- Email: [email protected]
Pathogenesis of diabetic retinopathy—The missing link?
Medical Hypotheses I
1
M,&d&po&ma(1993)41,205-210 lDLmgmmclmupuKLtdl993
Pathogenesis of Diabetic Retinopathy-The Link?
Missing
D. VERMA Blackbum
Royal Infirmary, Bolton Road, Blackburn,
BB2 3LR, UK
Abstract-Release of angiogenic factors in response to the ischaemic retina is currently the most favoured hypothesis for the pathogenesis of proliferative diabetic retinopathy. Reducing the stimulus for angiogenesis by destroying the ischaemic retina also forms the basis of the most effective treatment of diabetic retinopathy by photocoagulation. Though ischaemia is undoubtedly important for neovascularization, there is recent evidence which cast doubts on ischemia being the sole mechanism for diabetic retinopathy. Many clinical observations viz. the protective effects of glaucoma, myopia, unilateral carotid stenosis on diabetic retinopathy; and its worsening after cataract extraction are not adequately explained by the present hypothesis. Moreover, the recent in vitro culture studies on retinal pigment epithelial cells have suggested an a!ternative explanation for the effectiveness of photocoagulation in proliferative diabetic retinopathy. Furthermore, hyperglycemia has been strongly correlated with the incidence and progression of diabetic retinopathy, but has only been indirectly indicted in its pathogenesis. These facts can be integrated into a more plausible hypothesis for the pathogenesis of diabetic retinopathy. It is hypothesized that a relative reduction in intra-ocular pressure caused by presistent or intermittent hyperglycemia may be the missing link that induces certain morphological changes in the retinal pigment epithelium. These changes, in addition to the ischaemic retina, may be important for the pathogenesis of diabetic retinopathy. Such a hypothesis also explains most of the hitherto unexplained features of diabetic retinopathy.
Proliferative diabetic retinopathy is thought to result from the angiogenic factors liberated in response to the ischaemic retina (l-4). Selective destruction of the ischaemic area has been hypothesised to be the mechanism of improvement in retinopathy by photocoagulation (5). However some clinical observations and newer experimental revelations cast doubts on retinal ischaemia as the sole mechanism for the pathogenesis of diabetic retinopathy. Alternative explanations for the effectiveness of photocoagulation have
also been forthcoming. It is the purpose of this review to address this evidence and integrate available information into a more plausible hypothesis. Alternative explanations for the effectiveness of photocoagulation The original observation of Aiello et al (6) that eyes with a great deal of retinochoroidal scarring from trauma or inflammatory disease had markedly re-
Date received 12 May 1992 Date accepted 13 July 1992
205
MEDICAL HYPOTHESES
206 duced prevalence and severity of diabetic retinopathy, led to the use of ruby laser photocoagulation in the treatment of diabetic retinopathy. This effect has been thought to result from destruction of ischaemic retina and consequent decrease in oxygen demand (5). But recently Glaser and co-workers (7) have suggested an alternative explanation for the inhibitory effects of photocoagulation or retinochoroidal scarring on neovasculatization. In their in-vitro culture studies they have demonstrated that retinal pigment epithelium releases an inhibitor of neovascularization. The release of an inhibitor is dependent on retinal pigment epithelial cell density and associated cell morphology. They suggested that morphological alterations in the highly ordered retinal pigment epithelial monolayer that occur after photocoagulation or retinochoroidal scarring may result in enhancement of inhibitor release. In fact, there is also experimental evidence that only mild retinal lesions which do not destroy the overlying retina may induce some beneficial effects of photocoagulation (8). Thus retinal pigment epithelium may play a more important role than has been hitherto envisaged. Evidence against retinal ischaemia as the only mechanism for retinal neovascularization The occurrence of disc new vessels in patients with retinitis pigmentosa without any diabetes, is somewhat surprising (9). Since rod and cone activity is responsible for the major consumption of oxygen in the retina (lo), it is unlikely that the metabolic requirement of dystrophic retina is high. Very interestingly, the only animal model of retino-vitreal neovascularization, which comes closest to proliferative diabetic retinopathy, also develops in rats with hereditary retinal dystrophy with profound degeneration of the retina (11). The need for oxygen in these retinae should have been even less than that of a retina that has undergone extensive destruction from photocoagulation. On the other hand, there are reports of many diabetic patients with extensive retinal ischaemia who do not produce new vessels (12). Unexplained features of diabetic retinopathy Effect of glaucoma
Previously Mooney (13) reported that glaucoma reduced the prevalence and severity of diabetic retinopathy in affected eyes. This has never been confirmed in a methodologically precise epidemiological study, although some clinical observations support the claim (14). It has been suggested that the beneficial effects could be the result of reduction in the oxygen
demand by the damage to the retinal ganglion cell or optic nerve fibre layer (14). Such an explanation seems to be incorrect because diabetic retinopathy worsens in the eyes after filtering glaucoma operation (15). Therefore the protective effect of glaucoma must be related to the raised intraocular pressure only and not to the reduction in the oxygen demand. Myopia
Several years ago it was reported that myopia reduced the prevalence and severity of diabetic retinopathy (16). This observation has recently been confirmed in a large study by Rand et al (17). No adequate explanation is available for this observation. Unilateral carotid stenosis
Gay and Rosenbaum (18) reported that unilateral narrowing of internal carotid artery due to atherosclerosis protected the ipsilateral eye from developing diabetic retinopatby. No explanation has been offered for this intriguing phenomenon. Carotid narrowing may be expected to worsen the retinal ischemia rather than improve it. Cataract extraction
There have been many reports of progression of diabetic retinopathy after cataract extraction with or without intraocular lens implantation (19, 20). There is no adequate explanation other than a possible breakdown of blood-retinal barrier for some reason (19). Role of hyperglycemia Several large studies have now shown an unequivocal relationship of incidence and progression of diabetic rctinopathy to hyperglycemia (21-24). Retinopathy is seen in idiopathic as well as in secondary diabetes mellitus (25). There is also experimental evidence that hyperglycemia is the key factor in production of diabetic retinopatby changes (26). Such changes are also seen in experimental galactosemia (27). Despite such overwhelming data on the importance of hyperglycemia in diabetic retinopathy, the precise mechanism/s of pathogenesis of retinopathy, in relation to hyperglycemia, is uncertain. It has been shown that when retinal microvascular cells are grown invitro with a live fold excess of glucose over the physiological concentrations, protein synthesis is increased (28) but cell proliferation decreases substantially (28, 29). This may not be actually important
PATH’XENESIS
OF DIABETIC
RETINOPATHY
in diabetic patients because the pericytes and endothelial cells form a stable population with minimal or no cell division after fetal life (30). Alternatively, basement membrane thickening due to metabolism of glucose by the sorbitol pathway may be important. Such basement membrane thickening has been reported in galactosemic animals and is prevented by aldose-reductase inhibitors (31, 32). However, many clinical trials using aldose-reductase inhibitors for 3 years or more in early retinopathy have not shown any beneficial effects (33). Hence hyperglycemia, though clearly related to diabetic retinopathy, has not been directly indicted in its pathogenesis. New hypothesis It is well known that hyperosmolar agents iower
the intraocular pressure by producing an increase in the blood osmolarity (34). Hyperglycemia in patients with diabetes mellitus is also likely to result in a relative reduction in the intraocular pressure by raising the blood osmolarity. In fact, Larsen and Paulson (35) in a long forgotten study published in 1962 did demonstrate a decrease in intraocular pressure which correlated with high blood glucose in patients with diabetes (35). This relative ocular hypotension induced by hyperglycemia may cause morphological and functional changes in the pigment epithelium. Indeed, morphological alterations are found in the pigment epithelium early in experimental diabetes mellitus in rats (36). These authors report that there are increased infoldings of the plasma membrane at the basal surface of retinal pigment epithelial cells adjacent to the chorio-capillaries, in streptozotocin-induced diabetic rats. They also suggest that these infoldings increase the area of basal surface of retinal pigment epithelium, to make it more effective in active (or passive) transport of molecules. In fact, the basal membrane of the retinal pigment epithelium has also been found to be the principle site of small hyperosmotic loads in chick retinal pigment epithehum (37). It is hypothesised that the relative reduction in intra-ocular pressure might be responsible for such increased infoldings of basal membrane in diabetics early in the course of the disease. This in turn may cause the functional alterations in the pigment epithelium, which may be one of the factors responsible for the break down of the blood-retinal barrier, probably the earliest functional lesions thought to occur in diabetes mellitus (38-40). It has in fact been suggested that a leak in the pigment epithelium might be the site of breakdown of blood-retinal bar-
207 rier, early in the course of diabetes (38, 4.1:).These increased infoldings of basal membrane which facilitate abnormal transport across the pigment epithelium, may also be responsible for the transport of certain factors which induce proliferation of retinal vessels. The recent interest in the studies on the role of Insulin-like growth factors in the pathogenesis of diabetic retinopathy (42-45) also becomes more relevant. Merimee (42) has shown that patients with diabetes who have rapidly accelerating retinal neovascularisation and proliferation have high serum insulinlike growth factor-I values. Furthermore, it has been demonstrated that the concentration of insulin-like growth factor-1 in the vitreous correlates positively with serum concentrations in patients with diabetes but not in the normal subjects (45). This observation indicates that transport of these growth factors into the vitreous is somehow facilitated in patients with diabetes. Even though these factors may be produced in response to the ischaemic retina, the increased basal infoldings may have a significant role to play in their transport into the retina and vitreous. The relative hypotony caused by hyperglycemia may also help in accentuating most of the other features of diabetic retinopathy. Perfusion pressure of retinal vessels is normally the difference between the intra-vascular and the intraocular pressures. With the reduction in the intraocular pressure there would be a relative increase in the perfusion pressure in diabetes. This could worsen the venous dilatation and marked increase in the retinal blood how in diabetes noticed clinically (46) and also in experimental hyperglycemia (47). These vascular phenomena may be further enhanced by the loss of autoregulation that is known to occur in diabetes (48). With the relative rise of perfusion pressure, increased exudation into the retina especially through the diseased vessels in diabetics, explains the early exudative features of diabetic retinopatby. The increased perfusion pressure may also contribute to the production of micro aneurysms by exerting pressure on the vessel wall weakened by the loss of pericytes, known to occur in the diabetics (49). It is interesting to note that even though loss of pericytes is seen in other vessels in the body, microaneurysms have not been reported from any organ other than the eye with the exception of heart muscles (50) which again are subjected to the pressure changes in a closed chamber. Moreover the other conditions causing microaneurysms in the retina, like macroglobulinemia (51) and leukemia (52), also show increased osmolarity of the blood.
208
MEDICAL HYPOTHESES
Diabetes
mellitus
Microaneurysms Venous dilatation
Il~lntivc perfusion
r4y 0 p i a
P~iotocongulatio~~
(-) r
Irlceased tratlsport like proliferative
of JGI-1 factors
\
11rrakdowr1 t)lrlrld-rctirlnl L-ier I bar JI
of
Proliferative changes
DIABETIC
Fig See text. (+) Factors worsening like growth factor-l.
L/
J
1 Backgrourld changes
u/ Kllrlnt-tlr I in 1 damage
(-) V
Y
rise i,r pressure
\1 RETIHOPA’I’HY
the changes of diabetic retinopathy.
Explanation of hitherto unexplained features by the new hypothesis Effects of glaucoma
It may be speculated that the raised intra-ocular pressure in a diabetic patient may counteract the effects of relative hypotension induced by hyperglycemia and hence protect the eye against diabetic retinopathy. Myopia
Raised intra-ocular pressure has been incriminated in the etiology of myopia (53, 54) which could in turn
(-) Factors protecting
against diabetic retinopathy,
IGI-I
: Insulin
explain its protective effect in diabetic retinopathy. It is interesting to note that the retinal pigment epithelial cells are flatter and larger in myopia (55). Such cells may resist the morphological changes of increased basal infoldings induced by the relative hypotony.
Unilateral carotid stenosis The protective effect of unilateral narrowing of internal carotid artery on the ipsilateral eye can be explained by the fact that the intravascular pressure would be low in such cases. This would counteract the effect of low intra-ocular pressure and the per-
209
PATHOGENESIS OF DIABETK REI-INOPATKY
fusion pressure would then remain relatively normal, thereby reducing the chances of exudative phenomena of diabetic retinopathy.
features of diabetic retinopathy, though like any other hypothesis it remains to be proven. Acknowledgement
Cataract extraction
Worsening of diabetic retinopathy after cataract extraction can also be due to the slight reduction in intra-ocular pressure that is known to occur after cataract extraction (56). This would accentuate or mimic the effects of relative hypotony induced by hyperglycemia. The same mechanism will also explain the worsening of retinopathy after filtering glaucoma surgery (15). Explanation of some other features by the new hypothesis Hypertension
Direct correlation of hypertension with worsening of diabetic retinopathy has been well documented (57, 21). This may be related to the relative increase in the perfusion pressure. In fact such a mechanism has recently been suggested (58), and fits in well with the new hypothesis. Effects of photocoagulation or retino-choroidal scarring The beneficial effects of photocoagulation or retinochoroidal scarring can also be explained by the new hypothesis. The morphological changes in the pigment epithelium, induced by stretching of the normal retina in between the scars, may reduce the basal infoldings induced by low intraocular pressure. Such reduction in basal infoldings may restore the competence of the blood-retinal barrier. In fact, restoration of the blood-retinal barrier has been recognised following photocoagulation (59). The destruction of pigment epithelial cells which were responsible for the abnormal transport of growth factors may also play a significant role in the regression of new vessels. It is noteworthy that the beneficial effects of grid laser in diffuse macular oedema has already been explained by a similar mechanism (60). Thus a relative reduction in intraocular pressure induced by hyperglycemia may be the missing link (Fig.) that enhances the various pathological processes in the pathogenesis of diabetic retinopathy. The pigment epithelium, in addition to the ischaemic retina, may be the key to this complex process. This hypothesis explains most of the hitherto unexplained
I amthankful to Dr Shyam Sunder Kothari MD, DM (Cardio) for his constant encouragement and review of the manuscrip.
References 1. Michaelson I C. Mode of development of the vascular systan of the retina, with some observations on its significance for certain retinal disease. Trans Ophthalmd Sot UK 1948, 68: 137-180. Wise G N. Retinal neovascularization. Tram Am Ophthalmol Sot 1956; 54: 129626. Henkind P Ocular neovascularization. Am J Ophthalmol 1978; 85: 287-301. Patz A I. Studies on retinal neovascularization. Invest Ophthahnol Vis Sc 1980; 19: 1133-1138. Weiter J J, Zukennan R. The influence of photoreceptor RPE complex on the inner retina: an explanation for the beneficial effects of photocoagulation. Ophthalmology 1980, 87: 1133-1139. 6. Aiello L M, Beetham W P, Balodimos M C, Chazan B I, Bradley R F. Ruby laser photocoagulation in treatment of diabetic proliferative retincpathy: preliminary report. In: Goldberg M F, Fine S L, ers. Symposium on the treatment of diabetic retinopathy, US Public Health Service publication no. 1890. Washington DC: US Government printing office, 1968. I. Glaser B M, Campochiam P A, Davis J L, Jr. Sate M. Retinal pigment epitbelial cells release an inhibitor of neovascularization. Arch Ophthahnol 1985; 103: 1870-1875. 8. Marshall J, Clover G, Rothety S. Some new findings on retinal irradiation by krypton and argon lasers. Dot Ophthahnol Proc Ser 1984; 36: 21-37. 9. Linsenmeier R A. Effects of light and darkness on oxygen distribution and consumption in cat retina. J Gen Physol 1986; 88: 521-542. 10. Uliss A E, Gregor Z J. Bird A C. Retinitis pigmentosa and retinal neovascularization. Ophthalmology 1986; 93: 1599-1603. 11. Frank R N, Mancini M A. Presumed mtinovitreal neovascularization in dystrophic retinas of spontaneously hypertensive rats. Invest Ophtha&ol Vis Sci 1986; 27: 346355: 12. Verdaeuer J. le Clerco N. Holuieue J. Musalem R. Ncsmroliferative diabetic retit&&y w& significant capillary natperfusion. Graefes Arch Clin Exp Ophthalmol 1987; 225: 157-159. challenge. Br J Oph13. Mooney A J. Diabetic retinopathy-a thalmol 1963; 47: 513-520. 14. Frank R N. Etiological mechanism in diabetic retinopathy. In: Schachat A P, Murphy R B, Patz A, eds. Retina (~01-11) St Louis: The CV Mosby Co, 1989; 301-326. 15. Schcnherr U, Ruprecht K W, Lang G K. Unilateral progressive diabetic mtinopathy after filtering glaucoma operation. Khn Monatsbl Augenheilkd. 1988; 192: 44-46. 16. Jain I S, Luthra C L, Das T. Diabetic retinopathy and its relation to errors of refraction. Arch Ophthalmol 1%7; 77: 59-60. 17. Rand L I, Krolewski A S. Aiello C M, Warrant J H, Baker R S, Maki T. Multiple factors in the prediction of risk of proliferative diabetic retinopathy. N Engl J Med 1985; 313: 1433-1438. 18. Gay A J, Rosenbaum A C. Retinal artery pressure m assymettic diabetic retinopathy. Arch Ophthalmol 1%6; 75: 758-762. 19. Prasad P. Setna P H, Dunne J A. Accelerated ocular neovas-
210
20.
21.
22
23.
24.
25.
26.
27. 28.
29.
30. 31.
32.
33.
34. 35. 36.
31.
38.
MEDICAL HYl’OTHESEs
cularization in diabetes following posterior chamber lens implantation. Br J Ophthalmol 1999 74: 313-314. Viganelli M. Progression of diabetic retinopathy following cataract surgery. Klin Monastsbl Augenheilkd 1990, 196: 334-357. Janka H U, Warram J H, Rand L I. Krolewski A S. Risk factors for progression of background retinopathy in lcng-standing Insulin dependent diabetes mellitus. Diabetes 1989; 38: 460-464. Frank R N, Hoffman W H, Podger M J et al. Retinopathy in juvenile-onset diabetes of short duration. ophthalmology 1980; 87: l-9. Klein R, Klein B E K, Moss S E. Davis M D. DeMets D L. The Wisconsin Epidemiological Study of Diabetic Retinopathy II. Prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years. Arch Ophthalmol 1984; 102: 520-526. Palmerg P, Smith M, Waltman S et al. Natural history of retinopathy in insulin-dependant juvenile-anset diabetes. Ophthalmology 1981; 88: 613-618. Gullo L, Parenti M, Monti L, Pezzilli R. Barbara L. Diabetic retinopathy in chronic pancreatitis. Gastroenterology 1990; 98: 1577-1581. Engerman R, Bloodworth J M B Jr, Nelson S. Relationship of microvascular disease in diabetes to metabolic control. Diabetes 1977; 26: 760-769. Engerman R L, Kern T S. Experimental galactosemia produces diabetic-like retinopathy. Diabetes 1984; 33: 97-100. Li W, Shen S, Khatami M, Rockey J H. Stimulation of retinal capillary per&e protein and collagen synthesis in culture by high-glucose concentration. Diabetes 1984; 33: 785-789. King G L, Bumey S M. Modulation of growth in retinal capillary cells by insulin and hyperglycemia: possible roles in the development of early diabetic retinopathy. Invest Ophthalmol Vis Sci 1983; 24(suppl): 14. Engerman R L, Pfaffenbach D. Davis M D. Cell turnover of canillaries. Lab Invest 1967: 17: 738-743. Frank R N, Keim R J, Kennedy A, Frank K W. Galactose-induced retinal capillary basement membrane thickening: prevention by Sorbinol. Invest Ophthahnol Vis Sci 1983; 24: 1519-1524. Robinson W J Jr, Kador P F, Kinoshita J H. Retinal capillaries: basement membrane thickening by galactosemia prevented with aldose reductase inhibitor. Science 1983; 221: 1177-l 179. Kohner E M. Retinal ischaemia. In: Schachat A P, Murphy R E, Patz A, eds. Retina (vol II). St Louis: The CV Mosby Co, 1989: 89-97. Galin M A, Binkhorst R D, Kwitko M L. Ocular dehydration. Am J Ophthalmol 1%8: 66: 233. Larson H W, Poulsen J E. Intra-ocular tension and blood sugar fluctuations in diabetes. Acta Ophthalmol 1962; 40: 58&589. Grimes P A, Laties A M. Early morphological alteration of pigment epithelium in streptozotocin-induced diabetes: increased surface area of the basal cell membrane. Exp Eye Res 1980; 30: 631439. Shirao Y, Steinberg R H. Mechanism of effects of small hypcrosmotic gradient on the chick retinal pigment epithelium. Invest Ophthalmol Vis Sci 1987; 28: 2015-2025. Kirber W M, Nicholas C W, Grimes P A, Winegard A I, Laties A M. A pcrmiability defect of retinal pigment epithelium: oc cutrence in early streptozctocin diabetes. Arch Ophthahnol 1980; 98: 725-728.
39. Waltman S R, Santiago J, Krupin T, Singer P, Becker B. Blucher S. Vitreous fluorophotometry and blood sugar control in diabetes. Lancet 1979; 2: 1068. 40. Cuhna-Vaz J G, Fonseca J R, Abreu J F, Russ M A. A follow up study by vitreous fluorophotometty of early retinal involvement in diabetes. Am J Gphthalmol 1978; 86: 1185-1190. 41. Vinores S A, Campochiaro ‘PA. Ultrastructural localisation of blood-retinal barrier breakdown in diabetic and galactosunic rats. J Histochem Cytochem 1990. 38(9): 1341-1352. 42. Merimee T J, Zapf J. Froesch E R. Insulin-like growth factors studies in diabetes with or without retinopathy. N Engl J Med 1983; 309: 527-530. 43. Grant M, Russel B, Fitzgerald C, Merimee T J. Insulin-like growth factors in vitreous: studies in control and diabetic subjects with neovasculatizatiat. Diabetes 1986; 35: 416-Q.O. 44. Bassnett S, Beebe D C. Localization of insulin-like growth factor-l binding sites in embryonic chick eye. Invest Ophthalmol Vis Sci 1990; 31: 1637-1643. 45. Grant M B, Guay C, Marsh R. Insulin-like growth factor-1 stimulates proliferation, migration and plasminogen activator release by human retinal pigment epithelial cells. Gun Eye Res 1990; 9: 323-335. 46. Kohner E M, Hamilton A M. Saunders S J, Sutcliffe B A, Bulpitt C J. The retinal blood flow in diabetes. Diabetologia 1975; 11: 27-33. 47. Athencn A, Hill D W, Keen H, Young S, Edwards E J. The effect of acute hyperglycemia on the retinal circulation of the normal cat. Diabetologia 1980; 18: 233-273. 48. Sinclair S H. Gmnwald J E, Riva C E, Braunstein S N, Nicholas C W, Schwartz S S. Retinal vascular autoregulation in diabetes mellitus. Ophthalmology 1982; 89: 748-750. 49. Addison D J, Gamer A, Ashton N. Degeneration of intramural pcricytes in diabetic retinopathy. Br Med J 1970; 1: 264-266. 50. Factor S M, Okun E M, Minase T. Capillary microaneurysms in human diabetic heart. N Eng J Med 1980; 302: 384-388. 51. Ashton N, Kok D’A, Foulds W S. Ocular pathology in macroaloblunaemia. J Path01 Bacterial 1963: 86: 453461. 52. Duke YR, Wilkinson C P, Sigelman S. Retinal microaneurysm in leukaemia. Br J Gphthahnol 1968; 52: 368-374. 53. Maurice D M. Mushin A S. Production of myopia in rabbits by raised body-temperature and increased intra-ocular pressure. Lancet 1966; 2: 1160-1162. 54. Greene P R. Mechanical considerations in myopia: relative effects of accommodation, convergence, intra-ocular pressure and the extra-ocular muscles. Am J Gptom Physiol Opt 1980; 57: 902-914. 55. Greene J L, Rabb M F. Degeneration of Bruch’s membrane and retinal epithelium. Int Ophthalmol Clin 1981; 21(3): 27-50. 56. Allen Store-Paulsen. Surgically Induced Astigmatism. Eur J Implant Ref Surg 1991; 3: 249-253. 57. Knowler W C, Bennett P H, Ballintine E J. Increased incidence of retinopathy in diabetes with elevated blood pressure: a six year follow up study in Pima Indians. N Engl J Mcd 1980; 302: 645-650. 58. Merimee T J. Mechanisms of Diseaes. Diabetic retinopathy -a synthesis of perspectives. N Engl J Med 1990, 322: 978-983. 59. Wallow I H. Repair of the pigment epithelial barrier following photocoagulation. Arch Ophthalmol 1984; 102: 126-135. 60. Bresnick G H. Diabetic maculopathy: a critical review highlighting diffuse macular oedema. Ophthalmology 1983; 90: 1301-1317.
1
M,&d&po&ma(1993)41,205-210 lDLmgmmclmupuKLtdl993
Pathogenesis of Diabetic Retinopathy-The Link?
Missing
D. VERMA Blackbum
Royal Infirmary, Bolton Road, Blackburn,
BB2 3LR, UK
Abstract-Release of angiogenic factors in response to the ischaemic retina is currently the most favoured hypothesis for the pathogenesis of proliferative diabetic retinopathy. Reducing the stimulus for angiogenesis by destroying the ischaemic retina also forms the basis of the most effective treatment of diabetic retinopathy by photocoagulation. Though ischaemia is undoubtedly important for neovascularization, there is recent evidence which cast doubts on ischemia being the sole mechanism for diabetic retinopathy. Many clinical observations viz. the protective effects of glaucoma, myopia, unilateral carotid stenosis on diabetic retinopathy; and its worsening after cataract extraction are not adequately explained by the present hypothesis. Moreover, the recent in vitro culture studies on retinal pigment epithelial cells have suggested an a!ternative explanation for the effectiveness of photocoagulation in proliferative diabetic retinopathy. Furthermore, hyperglycemia has been strongly correlated with the incidence and progression of diabetic retinopathy, but has only been indirectly indicted in its pathogenesis. These facts can be integrated into a more plausible hypothesis for the pathogenesis of diabetic retinopathy. It is hypothesized that a relative reduction in intra-ocular pressure caused by presistent or intermittent hyperglycemia may be the missing link that induces certain morphological changes in the retinal pigment epithelium. These changes, in addition to the ischaemic retina, may be important for the pathogenesis of diabetic retinopathy. Such a hypothesis also explains most of the hitherto unexplained features of diabetic retinopathy.
Proliferative diabetic retinopathy is thought to result from the angiogenic factors liberated in response to the ischaemic retina (l-4). Selective destruction of the ischaemic area has been hypothesised to be the mechanism of improvement in retinopathy by photocoagulation (5). However some clinical observations and newer experimental revelations cast doubts on retinal ischaemia as the sole mechanism for the pathogenesis of diabetic retinopathy. Alternative explanations for the effectiveness of photocoagulation have
also been forthcoming. It is the purpose of this review to address this evidence and integrate available information into a more plausible hypothesis. Alternative explanations for the effectiveness of photocoagulation The original observation of Aiello et al (6) that eyes with a great deal of retinochoroidal scarring from trauma or inflammatory disease had markedly re-
Date received 12 May 1992 Date accepted 13 July 1992
205
MEDICAL HYPOTHESES
206 duced prevalence and severity of diabetic retinopathy, led to the use of ruby laser photocoagulation in the treatment of diabetic retinopathy. This effect has been thought to result from destruction of ischaemic retina and consequent decrease in oxygen demand (5). But recently Glaser and co-workers (7) have suggested an alternative explanation for the inhibitory effects of photocoagulation or retinochoroidal scarring on neovasculatization. In their in-vitro culture studies they have demonstrated that retinal pigment epithelium releases an inhibitor of neovascularization. The release of an inhibitor is dependent on retinal pigment epithelial cell density and associated cell morphology. They suggested that morphological alterations in the highly ordered retinal pigment epithelial monolayer that occur after photocoagulation or retinochoroidal scarring may result in enhancement of inhibitor release. In fact, there is also experimental evidence that only mild retinal lesions which do not destroy the overlying retina may induce some beneficial effects of photocoagulation (8). Thus retinal pigment epithelium may play a more important role than has been hitherto envisaged. Evidence against retinal ischaemia as the only mechanism for retinal neovascularization The occurrence of disc new vessels in patients with retinitis pigmentosa without any diabetes, is somewhat surprising (9). Since rod and cone activity is responsible for the major consumption of oxygen in the retina (lo), it is unlikely that the metabolic requirement of dystrophic retina is high. Very interestingly, the only animal model of retino-vitreal neovascularization, which comes closest to proliferative diabetic retinopathy, also develops in rats with hereditary retinal dystrophy with profound degeneration of the retina (11). The need for oxygen in these retinae should have been even less than that of a retina that has undergone extensive destruction from photocoagulation. On the other hand, there are reports of many diabetic patients with extensive retinal ischaemia who do not produce new vessels (12). Unexplained features of diabetic retinopathy Effect of glaucoma
Previously Mooney (13) reported that glaucoma reduced the prevalence and severity of diabetic retinopathy in affected eyes. This has never been confirmed in a methodologically precise epidemiological study, although some clinical observations support the claim (14). It has been suggested that the beneficial effects could be the result of reduction in the oxygen
demand by the damage to the retinal ganglion cell or optic nerve fibre layer (14). Such an explanation seems to be incorrect because diabetic retinopathy worsens in the eyes after filtering glaucoma operation (15). Therefore the protective effect of glaucoma must be related to the raised intraocular pressure only and not to the reduction in the oxygen demand. Myopia
Several years ago it was reported that myopia reduced the prevalence and severity of diabetic retinopathy (16). This observation has recently been confirmed in a large study by Rand et al (17). No adequate explanation is available for this observation. Unilateral carotid stenosis
Gay and Rosenbaum (18) reported that unilateral narrowing of internal carotid artery due to atherosclerosis protected the ipsilateral eye from developing diabetic retinopatby. No explanation has been offered for this intriguing phenomenon. Carotid narrowing may be expected to worsen the retinal ischemia rather than improve it. Cataract extraction
There have been many reports of progression of diabetic retinopathy after cataract extraction with or without intraocular lens implantation (19, 20). There is no adequate explanation other than a possible breakdown of blood-retinal barrier for some reason (19). Role of hyperglycemia Several large studies have now shown an unequivocal relationship of incidence and progression of diabetic rctinopathy to hyperglycemia (21-24). Retinopathy is seen in idiopathic as well as in secondary diabetes mellitus (25). There is also experimental evidence that hyperglycemia is the key factor in production of diabetic retinopatby changes (26). Such changes are also seen in experimental galactosemia (27). Despite such overwhelming data on the importance of hyperglycemia in diabetic retinopathy, the precise mechanism/s of pathogenesis of retinopathy, in relation to hyperglycemia, is uncertain. It has been shown that when retinal microvascular cells are grown invitro with a live fold excess of glucose over the physiological concentrations, protein synthesis is increased (28) but cell proliferation decreases substantially (28, 29). This may not be actually important
PATH’XENESIS
OF DIABETIC
RETINOPATHY
in diabetic patients because the pericytes and endothelial cells form a stable population with minimal or no cell division after fetal life (30). Alternatively, basement membrane thickening due to metabolism of glucose by the sorbitol pathway may be important. Such basement membrane thickening has been reported in galactosemic animals and is prevented by aldose-reductase inhibitors (31, 32). However, many clinical trials using aldose-reductase inhibitors for 3 years or more in early retinopathy have not shown any beneficial effects (33). Hence hyperglycemia, though clearly related to diabetic retinopathy, has not been directly indicted in its pathogenesis. New hypothesis It is well known that hyperosmolar agents iower
the intraocular pressure by producing an increase in the blood osmolarity (34). Hyperglycemia in patients with diabetes mellitus is also likely to result in a relative reduction in the intraocular pressure by raising the blood osmolarity. In fact, Larsen and Paulson (35) in a long forgotten study published in 1962 did demonstrate a decrease in intraocular pressure which correlated with high blood glucose in patients with diabetes (35). This relative ocular hypotension induced by hyperglycemia may cause morphological and functional changes in the pigment epithelium. Indeed, morphological alterations are found in the pigment epithelium early in experimental diabetes mellitus in rats (36). These authors report that there are increased infoldings of the plasma membrane at the basal surface of retinal pigment epithelial cells adjacent to the chorio-capillaries, in streptozotocin-induced diabetic rats. They also suggest that these infoldings increase the area of basal surface of retinal pigment epithelium, to make it more effective in active (or passive) transport of molecules. In fact, the basal membrane of the retinal pigment epithelium has also been found to be the principle site of small hyperosmotic loads in chick retinal pigment epithehum (37). It is hypothesised that the relative reduction in intra-ocular pressure might be responsible for such increased infoldings of basal membrane in diabetics early in the course of the disease. This in turn may cause the functional alterations in the pigment epithelium, which may be one of the factors responsible for the break down of the blood-retinal barrier, probably the earliest functional lesions thought to occur in diabetes mellitus (38-40). It has in fact been suggested that a leak in the pigment epithelium might be the site of breakdown of blood-retinal bar-
207 rier, early in the course of diabetes (38, 4.1:).These increased infoldings of basal membrane which facilitate abnormal transport across the pigment epithelium, may also be responsible for the transport of certain factors which induce proliferation of retinal vessels. The recent interest in the studies on the role of Insulin-like growth factors in the pathogenesis of diabetic retinopathy (42-45) also becomes more relevant. Merimee (42) has shown that patients with diabetes who have rapidly accelerating retinal neovascularisation and proliferation have high serum insulinlike growth factor-I values. Furthermore, it has been demonstrated that the concentration of insulin-like growth factor-1 in the vitreous correlates positively with serum concentrations in patients with diabetes but not in the normal subjects (45). This observation indicates that transport of these growth factors into the vitreous is somehow facilitated in patients with diabetes. Even though these factors may be produced in response to the ischaemic retina, the increased basal infoldings may have a significant role to play in their transport into the retina and vitreous. The relative hypotony caused by hyperglycemia may also help in accentuating most of the other features of diabetic retinopathy. Perfusion pressure of retinal vessels is normally the difference between the intra-vascular and the intraocular pressures. With the reduction in the intraocular pressure there would be a relative increase in the perfusion pressure in diabetes. This could worsen the venous dilatation and marked increase in the retinal blood how in diabetes noticed clinically (46) and also in experimental hyperglycemia (47). These vascular phenomena may be further enhanced by the loss of autoregulation that is known to occur in diabetes (48). With the relative rise of perfusion pressure, increased exudation into the retina especially through the diseased vessels in diabetics, explains the early exudative features of diabetic retinopatby. The increased perfusion pressure may also contribute to the production of micro aneurysms by exerting pressure on the vessel wall weakened by the loss of pericytes, known to occur in the diabetics (49). It is interesting to note that even though loss of pericytes is seen in other vessels in the body, microaneurysms have not been reported from any organ other than the eye with the exception of heart muscles (50) which again are subjected to the pressure changes in a closed chamber. Moreover the other conditions causing microaneurysms in the retina, like macroglobulinemia (51) and leukemia (52), also show increased osmolarity of the blood.
208
MEDICAL HYPOTHESES
Diabetes
mellitus
Microaneurysms Venous dilatation
Il~lntivc perfusion
r4y 0 p i a
P~iotocongulatio~~
(-) r
Irlceased tratlsport like proliferative
of JGI-1 factors
\
11rrakdowr1 t)lrlrld-rctirlnl L-ier I bar JI
of
Proliferative changes
DIABETIC
Fig See text. (+) Factors worsening like growth factor-l.
L/
J
1 Backgrourld changes
u/ Kllrlnt-tlr I in 1 damage
(-) V
Y
rise i,r pressure
\1 RETIHOPA’I’HY
the changes of diabetic retinopathy.
Explanation of hitherto unexplained features by the new hypothesis Effects of glaucoma
It may be speculated that the raised intra-ocular pressure in a diabetic patient may counteract the effects of relative hypotension induced by hyperglycemia and hence protect the eye against diabetic retinopathy. Myopia
Raised intra-ocular pressure has been incriminated in the etiology of myopia (53, 54) which could in turn
(-) Factors protecting
against diabetic retinopathy,
IGI-I
: Insulin
explain its protective effect in diabetic retinopathy. It is interesting to note that the retinal pigment epithelial cells are flatter and larger in myopia (55). Such cells may resist the morphological changes of increased basal infoldings induced by the relative hypotony.
Unilateral carotid stenosis The protective effect of unilateral narrowing of internal carotid artery on the ipsilateral eye can be explained by the fact that the intravascular pressure would be low in such cases. This would counteract the effect of low intra-ocular pressure and the per-
209
PATHOGENESIS OF DIABETK REI-INOPATKY
fusion pressure would then remain relatively normal, thereby reducing the chances of exudative phenomena of diabetic retinopathy.
features of diabetic retinopathy, though like any other hypothesis it remains to be proven. Acknowledgement
Cataract extraction
Worsening of diabetic retinopathy after cataract extraction can also be due to the slight reduction in intra-ocular pressure that is known to occur after cataract extraction (56). This would accentuate or mimic the effects of relative hypotony induced by hyperglycemia. The same mechanism will also explain the worsening of retinopathy after filtering glaucoma surgery (15). Explanation of some other features by the new hypothesis Hypertension
Direct correlation of hypertension with worsening of diabetic retinopathy has been well documented (57, 21). This may be related to the relative increase in the perfusion pressure. In fact such a mechanism has recently been suggested (58), and fits in well with the new hypothesis. Effects of photocoagulation or retino-choroidal scarring The beneficial effects of photocoagulation or retinochoroidal scarring can also be explained by the new hypothesis. The morphological changes in the pigment epithelium, induced by stretching of the normal retina in between the scars, may reduce the basal infoldings induced by low intraocular pressure. Such reduction in basal infoldings may restore the competence of the blood-retinal barrier. In fact, restoration of the blood-retinal barrier has been recognised following photocoagulation (59). The destruction of pigment epithelial cells which were responsible for the abnormal transport of growth factors may also play a significant role in the regression of new vessels. It is noteworthy that the beneficial effects of grid laser in diffuse macular oedema has already been explained by a similar mechanism (60). Thus a relative reduction in intraocular pressure induced by hyperglycemia may be the missing link (Fig.) that enhances the various pathological processes in the pathogenesis of diabetic retinopathy. The pigment epithelium, in addition to the ischaemic retina, may be the key to this complex process. This hypothesis explains most of the hitherto unexplained
I amthankful to Dr Shyam Sunder Kothari MD, DM (Cardio) for his constant encouragement and review of the manuscrip.
References 1. Michaelson I C. Mode of development of the vascular systan of the retina, with some observations on its significance for certain retinal disease. Trans Ophthalmd Sot UK 1948, 68: 137-180. Wise G N. Retinal neovascularization. Tram Am Ophthalmol Sot 1956; 54: 129626. Henkind P Ocular neovascularization. Am J Ophthalmol 1978; 85: 287-301. Patz A I. Studies on retinal neovascularization. Invest Ophthahnol Vis Sc 1980; 19: 1133-1138. Weiter J J, Zukennan R. The influence of photoreceptor RPE complex on the inner retina: an explanation for the beneficial effects of photocoagulation. Ophthalmology 1980, 87: 1133-1139. 6. Aiello L M, Beetham W P, Balodimos M C, Chazan B I, Bradley R F. Ruby laser photocoagulation in treatment of diabetic proliferative retincpathy: preliminary report. In: Goldberg M F, Fine S L, ers. Symposium on the treatment of diabetic retinopathy, US Public Health Service publication no. 1890. Washington DC: US Government printing office, 1968. I. Glaser B M, Campochiam P A, Davis J L, Jr. Sate M. Retinal pigment epitbelial cells release an inhibitor of neovascularization. Arch Ophthahnol 1985; 103: 1870-1875. 8. Marshall J, Clover G, Rothety S. Some new findings on retinal irradiation by krypton and argon lasers. Dot Ophthahnol Proc Ser 1984; 36: 21-37. 9. Linsenmeier R A. Effects of light and darkness on oxygen distribution and consumption in cat retina. J Gen Physol 1986; 88: 521-542. 10. Uliss A E, Gregor Z J. Bird A C. Retinitis pigmentosa and retinal neovascularization. Ophthalmology 1986; 93: 1599-1603. 11. Frank R N, Mancini M A. Presumed mtinovitreal neovascularization in dystrophic retinas of spontaneously hypertensive rats. Invest Ophtha&ol Vis Sci 1986; 27: 346355: 12. Verdaeuer J. le Clerco N. Holuieue J. Musalem R. Ncsmroliferative diabetic retit&&y w& significant capillary natperfusion. Graefes Arch Clin Exp Ophthalmol 1987; 225: 157-159. challenge. Br J Oph13. Mooney A J. Diabetic retinopathy-a thalmol 1963; 47: 513-520. 14. Frank R N. Etiological mechanism in diabetic retinopathy. In: Schachat A P, Murphy R B, Patz A, eds. Retina (~01-11) St Louis: The CV Mosby Co, 1989; 301-326. 15. Schcnherr U, Ruprecht K W, Lang G K. Unilateral progressive diabetic mtinopathy after filtering glaucoma operation. Khn Monatsbl Augenheilkd. 1988; 192: 44-46. 16. Jain I S, Luthra C L, Das T. Diabetic retinopathy and its relation to errors of refraction. Arch Ophthalmol 1%7; 77: 59-60. 17. Rand L I, Krolewski A S. Aiello C M, Warrant J H, Baker R S, Maki T. Multiple factors in the prediction of risk of proliferative diabetic retinopathy. N Engl J Med 1985; 313: 1433-1438. 18. Gay A J, Rosenbaum A C. Retinal artery pressure m assymettic diabetic retinopathy. Arch Ophthalmol 1%6; 75: 758-762. 19. Prasad P. Setna P H, Dunne J A. Accelerated ocular neovas-
210
20.
21.
22
23.
24.
25.
26.
27. 28.
29.
30. 31.
32.
33.
34. 35. 36.
31.
38.
MEDICAL HYl’OTHESEs
cularization in diabetes following posterior chamber lens implantation. Br J Ophthalmol 1999 74: 313-314. Viganelli M. Progression of diabetic retinopathy following cataract surgery. Klin Monastsbl Augenheilkd 1990, 196: 334-357. Janka H U, Warram J H, Rand L I. Krolewski A S. Risk factors for progression of background retinopathy in lcng-standing Insulin dependent diabetes mellitus. Diabetes 1989; 38: 460-464. Frank R N, Hoffman W H, Podger M J et al. Retinopathy in juvenile-onset diabetes of short duration. ophthalmology 1980; 87: l-9. Klein R, Klein B E K, Moss S E. Davis M D. DeMets D L. The Wisconsin Epidemiological Study of Diabetic Retinopathy II. Prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years. Arch Ophthalmol 1984; 102: 520-526. Palmerg P, Smith M, Waltman S et al. Natural history of retinopathy in insulin-dependant juvenile-anset diabetes. Ophthalmology 1981; 88: 613-618. Gullo L, Parenti M, Monti L, Pezzilli R. Barbara L. Diabetic retinopathy in chronic pancreatitis. Gastroenterology 1990; 98: 1577-1581. Engerman R, Bloodworth J M B Jr, Nelson S. Relationship of microvascular disease in diabetes to metabolic control. Diabetes 1977; 26: 760-769. Engerman R L, Kern T S. Experimental galactosemia produces diabetic-like retinopathy. Diabetes 1984; 33: 97-100. Li W, Shen S, Khatami M, Rockey J H. Stimulation of retinal capillary per&e protein and collagen synthesis in culture by high-glucose concentration. Diabetes 1984; 33: 785-789. King G L, Bumey S M. Modulation of growth in retinal capillary cells by insulin and hyperglycemia: possible roles in the development of early diabetic retinopathy. Invest Ophthalmol Vis Sci 1983; 24(suppl): 14. Engerman R L, Pfaffenbach D. Davis M D. Cell turnover of canillaries. Lab Invest 1967: 17: 738-743. Frank R N, Keim R J, Kennedy A, Frank K W. Galactose-induced retinal capillary basement membrane thickening: prevention by Sorbinol. Invest Ophthahnol Vis Sci 1983; 24: 1519-1524. Robinson W J Jr, Kador P F, Kinoshita J H. Retinal capillaries: basement membrane thickening by galactosemia prevented with aldose reductase inhibitor. Science 1983; 221: 1177-l 179. Kohner E M. Retinal ischaemia. In: Schachat A P, Murphy R E, Patz A, eds. Retina (vol II). St Louis: The CV Mosby Co, 1989: 89-97. Galin M A, Binkhorst R D, Kwitko M L. Ocular dehydration. Am J Ophthalmol 1%8: 66: 233. Larson H W, Poulsen J E. Intra-ocular tension and blood sugar fluctuations in diabetes. Acta Ophthalmol 1962; 40: 58&589. Grimes P A, Laties A M. Early morphological alteration of pigment epithelium in streptozotocin-induced diabetes: increased surface area of the basal cell membrane. Exp Eye Res 1980; 30: 631439. Shirao Y, Steinberg R H. Mechanism of effects of small hypcrosmotic gradient on the chick retinal pigment epithelium. Invest Ophthalmol Vis Sci 1987; 28: 2015-2025. Kirber W M, Nicholas C W, Grimes P A, Winegard A I, Laties A M. A pcrmiability defect of retinal pigment epithelium: oc cutrence in early streptozctocin diabetes. Arch Ophthahnol 1980; 98: 725-728.
39. Waltman S R, Santiago J, Krupin T, Singer P, Becker B. Blucher S. Vitreous fluorophotometry and blood sugar control in diabetes. Lancet 1979; 2: 1068. 40. Cuhna-Vaz J G, Fonseca J R, Abreu J F, Russ M A. A follow up study by vitreous fluorophotometty of early retinal involvement in diabetes. Am J Gphthalmol 1978; 86: 1185-1190. 41. Vinores S A, Campochiaro ‘PA. Ultrastructural localisation of blood-retinal barrier breakdown in diabetic and galactosunic rats. J Histochem Cytochem 1990. 38(9): 1341-1352. 42. Merimee T J, Zapf J. Froesch E R. Insulin-like growth factors studies in diabetes with or without retinopathy. N Engl J Med 1983; 309: 527-530. 43. Grant M, Russel B, Fitzgerald C, Merimee T J. Insulin-like growth factors in vitreous: studies in control and diabetic subjects with neovasculatizatiat. Diabetes 1986; 35: 416-Q.O. 44. Bassnett S, Beebe D C. Localization of insulin-like growth factor-l binding sites in embryonic chick eye. Invest Ophthalmol Vis Sci 1990; 31: 1637-1643. 45. Grant M B, Guay C, Marsh R. Insulin-like growth factor-1 stimulates proliferation, migration and plasminogen activator release by human retinal pigment epithelial cells. Gun Eye Res 1990; 9: 323-335. 46. Kohner E M, Hamilton A M. Saunders S J, Sutcliffe B A, Bulpitt C J. The retinal blood flow in diabetes. Diabetologia 1975; 11: 27-33. 47. Athencn A, Hill D W, Keen H, Young S, Edwards E J. The effect of acute hyperglycemia on the retinal circulation of the normal cat. Diabetologia 1980; 18: 233-273. 48. Sinclair S H. Gmnwald J E, Riva C E, Braunstein S N, Nicholas C W, Schwartz S S. Retinal vascular autoregulation in diabetes mellitus. Ophthalmology 1982; 89: 748-750. 49. Addison D J, Gamer A, Ashton N. Degeneration of intramural pcricytes in diabetic retinopathy. Br Med J 1970; 1: 264-266. 50. Factor S M, Okun E M, Minase T. Capillary microaneurysms in human diabetic heart. N Eng J Med 1980; 302: 384-388. 51. Ashton N, Kok D’A, Foulds W S. Ocular pathology in macroaloblunaemia. J Path01 Bacterial 1963: 86: 453461. 52. Duke YR, Wilkinson C P, Sigelman S. Retinal microaneurysm in leukaemia. Br J Gphthahnol 1968; 52: 368-374. 53. Maurice D M. Mushin A S. Production of myopia in rabbits by raised body-temperature and increased intra-ocular pressure. Lancet 1966; 2: 1160-1162. 54. Greene P R. Mechanical considerations in myopia: relative effects of accommodation, convergence, intra-ocular pressure and the extra-ocular muscles. Am J Gptom Physiol Opt 1980; 57: 902-914. 55. Greene J L, Rabb M F. Degeneration of Bruch’s membrane and retinal epithelium. Int Ophthalmol Clin 1981; 21(3): 27-50. 56. Allen Store-Paulsen. Surgically Induced Astigmatism. Eur J Implant Ref Surg 1991; 3: 249-253. 57. Knowler W C, Bennett P H, Ballintine E J. Increased incidence of retinopathy in diabetes with elevated blood pressure: a six year follow up study in Pima Indians. N Engl J Mcd 1980; 302: 645-650. 58. Merimee T J. Mechanisms of Diseaes. Diabetic retinopathy -a synthesis of perspectives. N Engl J Med 1990, 322: 978-983. 59. Wallow I H. Repair of the pigment epithelial barrier following photocoagulation. Arch Ophthalmol 1984; 102: 126-135. 60. Bresnick G H. Diabetic maculopathy: a critical review highlighting diffuse macular oedema. Ophthalmology 1983; 90: 1301-1317.